Sraffa, in his archives in the 1940s and 1950s, is quite appreciative of Karl Marx's analysis of capitalism. This appreciation contrasts with the opinion embodied in the label 'neo-Ricardian', which Bob Rowthorn invented.
I know about the passages below in the Sraffa archives from Riccardo Bellofiore. The archivist, Jonathan Smith, has dated this entry from 1955-1959, late in the writing of Production of Commodities.
I do not want to focus on whether Marx or Sraffa are correct or not. I would want to work out a simple example. Besides, Sraffa seems not convinced of how to analyze the reduction in the working day, when starting at prices.
But I want to note that Sraffa is very much using Marxist concepts: vulgar economics, labor values, prices of production, surplus value, exploitation, and rates of exploitation. And the analysis is based on Marx. Surplus value comes from extending the working day past the point at which workers reproduce their labor power.
"Use of the Notion of Surplus Value
"The prolongation of the working day beyond the point at which the labourer would have produced just an equivalent for the value of his labour-power ..." (Cap., Engels transl. p. 518) cp p. 539 [Chapter Sixteen: Absolute and Relative Surplus-Value]
Put it the other way round. If starting from capitalist society the working day is shortened till there is no surplus value left, this shortening must be equal for all: if it is, the prices of the commodities will change [owing to change in the rate of profits, which vanishes], but the wages will remain unchanged : if it is not, and the working day is reduced to the extent of the profits made in each industry, then prices would remain unchanged* after the shortening [for the number of (shorter) labor days, in industries having a high organic composition of capital, would increase in the same proportion as the fall of profits] but wages would be different.
[Footnote:] *(28.12.41) But profits would be different (after the reduction) in different industries!
[Marginal note:] c/p Letters of M and E 129-32 (letter of M. 2.8.62)
In other words, if we start from profits (as vulgar economy does) we reach the conclusion that the rate of exploitation is different in different industries, being higher in the more highly capitalised ones – which is not [and indeed contrary to] the fact. If we start from surplus value, which is equal in all industries, we get the correct measure of exploitation. The former conclusion is patent nonsense, and no view of exploitation could be based on it.
Note that the former (profits) goes with a theory of prices, the latter, of value (as defined below).
12.11.40 [Price is an exchange ratio which equalises rates of profit on capitals. Value is an exchange ratio which equalises rates of surplus-value on labour. If commodities exchanged at their values, profits would be different for different capitals, and capitals would move: therefore, this competition of capitals causes them to exchange at their prices.
The question is: are the rates of exploitation different? and if so why doesn’t labor move, and restore values and equality of rates of surplus value?]
The starting point is "the prolongation of the working day beyond the point at which the labourer would have produced just an equivalent for the value of his labour power" (Cap., Engels Tr. 518)
This point cannot be determined without reference to the value of the product (unless the labourer produces himself all the commodities he consumes).
But the point varies if we take value and if we take price.
Now, we are comparing the actual state with a hypothetical one in which only the necessary labour is performed.
In the actual state commodities are exchanged at their prices, whilst in the hypothetical state (where there would be nothing to be paid out in profits) at their values.
Which scale should we adopt for both states, in comparing them? It may be said: neither – each state has its own scale and that only is appropriate to it. No comparison can be made directly between the two extreme states. [marginal note says, "wrong, see p. 56.] We cannot imagine to move gradually from the actual state, shortening the working day; as we start from the actual state, we use its own scale, i.e. prices, in determining the ultimate goal towards which we move [and that will give different reductions for different branches of industry; but as we pass to successive other states, with shorter and shorter working days, the scale to be used changes, and prices move nearer (as the rate of profits is reduced) to values – so does the "point" aimed at change; until, on the threshold of the state in which only the necessary labour is performed, the prices will practically coincide with values, and the point aimed at with that determined by the scale of values, i.e. all labourers will have had their hours reduced in the same proportion. [The converse is true: if starting from the hypothetical state we prolong the working day by this method, we reach the actual state, having prolonged it for all labourers proportionally/equally, but through the change in prices having raise the profits in each branch proportionally with its capital]
Note that if we had adopted straightway values, and made the comparison between the two extreme cases, we should have obtained the same, correct, result. But if we had adopted prices, and made that comparison, it would have led us astray: the 'point' indicated by prices [i.e. different reductions in different industries] would have been false when the hyp. state was reached – for on the basis of values some labourers would be working more, and some less, than the necessary hours.
The imaginary process (described above, p. 3 bottom) of gradually shortening the working day, on the basis of the prices appropriate to each intermediate point, and therefore in different proportions for different industries, requires further consideration. As it stands, it is only correct at the wo extreme points [or rather only at the final point], but false at all intermediate ones: for, e.g., on the first step, when the rate of profit is reduced from 6 to 5%, the day of every worker must be reduced in one and the same proportion, and not in different proportions: it is clear that the latter method would give immediately absurd results.
In fact, this shows that the way in which I have argued the point on p. 1 is wrong (too weak). The objection of the vulgar economist is that the surplus produced in each industry (or firm) is measured by its profits. If he agreed to call it exploitation he would say that this is higher (absorbs a larger proportion of the working day) in the industries having more capital per worker. Therefore, he would have to conclude that, if exploitation has to be reduced in the different industries in such proportions as would maintain the rate of profit equal between them, at the lower level, this would require a larger reduction of the working day in the more capitalised industries. It can be shown, by a numerical example, that this is not the case. That on the contrary if the was reduced equally in all industries, the rate of profits would also be reduced equally. This result is made possible by a simultaneous change in prices – those of highly capitalised industries falling (when the rate of profits falls) relatively to the other prices. So that the larger fall in the surplus of such industries has two sources: a) the reduction in working way (common to other industries), (b) the fall in the relative price of their product (peculiar to the highly cap. industries)
29.12.41 [A third source, working in the same direction, would at first sight appear to be the increased depreciation allowances for capital, as the rate of interest falls. However this is a delusion: the 10 loom case shows that it is constant. That is to say, it is constant if real capital has to be maintained intact, though allowing value of semi-used capital to fall as rate of interest falls - and this is the relevant case. Its rise only if money value has to maintained as originally, in case
29.12.41 The previous paragraph is misleading. There is a third source, even if depreciation allowances are regarded as constant. For the value (or rather, in M's sense, the "price") of capital goods falls with the fall in the rate of interest. Therefore, when the rate of s.v. falls, the profits of industries with a large amount of capital per man fall still more owing to this third source - even if the capitals are no more (but no less!) "durable" than in other industries (if they are more durable in the highly capitalised industries than in others, this is a fourth source - for with fall in interest the more durable capitals falls more than that of the less durable ones).
[The whole subject of the "measure of capital" requires investigation in this connection. It has a striking similarity to the contradiction of "prices" and "value" of commodities, and it also depends on the equalisation of the rate of profits. One should start from the "value" of capitals (i.e. quantity of labour necessary to construct them) and see how the requirement of equal rates of profits leads to "prices" of capitals different from their "values".]
30.12.41 This business of 3 or 4 sources is wrong. There are only two sources: a) the reduction of working time in the industry, which reduces the quantity of goods produced; b) The reduction in the price at which the product is sold. The fall in the value of capital of certain industries along with the fall in the general rate of profits and other possible causes, contribute to source b, but don't add anything besides b.
[N.B. The fact that the value of capital (and therefore its "quantity" or magnitude) varies with the rate of profits (and generally cannot be known without knowing prices and rate of profits) makes nonsense of many cornerstones: 1) "Sacrifice of waiting", but how if they don’t know what they are abstaining from? 2) rate of interest, or marg. prod. of cap., as criterion for distribution of resources; but how, if the same resource (in "value") becomes larger or smaller (in "price") according as it used in one way or another?
5.1.42 Those who regard Marx's transition from values to prices, by the necessity of equalising the rate of profit, as a trick, should say the same of Ricardo's (and the whole marginal school) method of determining cost of production by considering only that on the marginal land, by the necessity of equalising the price of all bushels of corn, on whichever land they may be procured. Cannan does so (Rev. of Ec. Theory, p. 178): Ricardo 'did the trick by little more than an arbitrary exercise of the right to define terms ..." -- Piero Sraffa D3.12.46/57r – 63r
Nothing like the above is in Sraffa's book. Connections to Marx are less apparent, although some reviewers perceived them. Counterfactual reasoning is mostly eschewed. The length of the working day is not discussed, but taken as given.
Sraffa does not seem very confident about whether he should start with value or prices and how he should proceed if he adopts the latter. He does see the importance of what was later called price Wicksell effects. I want to note that the next pages in the archive are a draft of the chapter on land in Sraffa's book.
By the way, Ian Steedman has a chapter towards the end of Marx after Sraffa illustrating the analysis of the length of the working day. Consistent with his general approach, he uses data on physical quantity flows and does not take the point at which prices are values and labor is not exploited as a reference point.
Posts
No comments:
Post a Comment